M-theory and the birth of the Universe - Frans R. Klinkhamer - KIT ITP

 
CONTINUE READING
M-theory and the birth of the Universe - Frans R. Klinkhamer - KIT ITP
XXVIIth Cracow Epiphany Conference                                        January 7-10, 2021
on the Future of Particle Physics

     M-theory and the birth of the Universe
                      Frans R. Klinkhamer
                     Institute for Theoretical Physics,
                  Karlsruhe Institute of Technology (KIT),
                        76128 Karlsruhe, Germany
                     Email: frans.klinkhamer@kit.edu

in memory of Martinus J. G. Veltman (1931-2021)

                                                   Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 1
0. Introduction

Two preliminary remarks:

    The Future of Particle Physics may very well be tied to that of
     Gravitation and Cosmology.
    The present talk has no definitive answers, only a few suggestive
     results (the ermine on da Vinci’s painting appears to know it all, but does not tell us...).

                                                                 [Czartoryski Museum, Kraków, PL]

                                                              Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 2
0. Introduction

Outline of main talk:
 1. Standard Friedmann cosmology
 2. Regularized big bang
 3. New phase from M-theory
 4. Conclusions
 5. References

Technical details:
 A. Extraction of the spacetime points
 B. Extraction of the spacetime metric
 C. Various emergent spacetimes
 D. More on the Lorentzian signature

                                         Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 3
1. Standard Friedmann cosmology

The Einstein gravitational field equation of general relativity (GR)
reads [1]:
                             1                  (SM)
                     Rµν − gµν R = −8πG Tµν          ,                                                     (1)
                             2
                                                 (SM)
with Rµν the Ricci tensor, R the Ricci scalar, Tµν    the energy-momentum
tensor of the matter described by the Standard Model (SM), and G the
Newton gravitational coupling constant. The spacetime indices µ, ν
run over {0, 1, 2, 3}.

For cosmology, the spatially flat Robertson–Walker (RW) metric is

                (RW)                      (RW)
            2                     µ   ν
       ds              ≡ gµν (x) dx dx           = −dt2 + a2 (t) δij dxi dxj ,                             (2)

with x0 = c t and c = 1. The spatial indices i, j run over {1, 2, 3}.

                                                         Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 4
1. Standard Friedmann cosmology

Considering a homogeneous perfect fluid with energy density ρM (t) and
pressure PM (t), we get the spatially flat Friedmann equations [1]:
                         2
                          ȧ      8πG
                               =        ρM ,                        (3a)
                          a         3
                                2
                        ä 1 ȧ
                           +            = −4πG PM ,                 (3b)
                        a 2 a
                                ȧ h           i
                        ρ̇M + 3     ρM + PM = 0 ,                   (3c)
                                a
                                        
                        PM = PM ρM ,                                (3d)

where the overdot stands for differentiation with respect to t and (3d)
corresponds to the equation-of-state (EOS) relation between pressure
and energy density of the fluid.

                                              Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 5
1. Standard Friedmann cosmology

For relativistic matter with constant EOS parameter wM ≡ PM /ρM = 1/3,
the Friedmann–Lemaître–Robertson–Walker (FLRW) solution is [1]
                (wM =1/3)      p
          a(t)             =     t/t0 ,              for t > 0 , (4a)
               FLRW
               (wM =1/3)
      ρM (t)               =   ρM 0 /a4 (t) ∝ 1/t2 ,              for t > 0 ,                     (4b)
               FLRW

where the cosmic scale factor has normalization a(t0 ) = 1 at t0 > 0.

This FLRW solution displays the big bang singularity for t → 0+ ,
                               lim+ a(t) = 0 ,                                                      (5)
                               t→0
with diverging curvature and energy density. But, at t = 0, the theory
(GR+SM) is no longer valid and we can ask what happens really?
Or, more precisely, how to describe the birth of the Universe?

                                                  Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 6
2. Regularized big bang

First, we set out to control the divergences by using a new Ansatz
for the “regularized” big bang [2]:
                  (reg-bb)                           (reg-bb)
              2                            µ    ν
         ds                  ≡   gµν (x) dx dx

                                     t2    2     2            i   j
                             =   − 2     dt  + a   (t) δ ij dx  dx  ,                                      (6a)
                                  t + b2

                       b2    >   0,   a2 (t) > 0 ,                                                         (6b)

                         t   ∈   (−∞, ∞) ,      xi ∈ (−∞, ∞) ,                                             (6c)

with x0 = c t and c = 1. The length scale b 6= 0 acts as regulator.

This metric gµν (x) is degenerate, with a vanishing determinant at t = 0.
Physically, the t = 0 slice corresponds to a spacetime defect.

                                                           Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 7
2. Regularized big bang

The standard Einstein equation (1) with the new metric Ansatz (6) and
a homogeneous perfect fluid gives modified spatially flat Friedmann
equations:
                   2
                        2
                  b      ȧ     8πG
              1+ 2            =        ρM ,                        (7a)
                  t      a        3
                                    !
                                      2
                  b2     ä 1 ȧ            b2 ȧ
              1+ 2          +             − 3 = −4πG PM ,          (7b)
                  t      a 2 a              t a

                      ȧ h        i
              ρ̇M + 3      ρM + PM = 0 ,                                                       (7c)
                      a
                             
              PM = PM ρM ,                                                                     (7d)

where the overdot stands again for differentiation with respect to t.

                                               Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 8
2. Regularized big bang

For constant EOS parameter wM = 1/3, the solution a(t) of (7) reads

                       (wM =1/3)           q                                    
                a(t)                   =   4
                                               t2   +   b2         t20   +   b2       ,                             (8)
                       (reg-bb)

which is perfectly smooth at t = 0 as long as b 6= 0. The figure below
compares this regularized solution (full curve) with the singular FLRW
solution (dashed curve).
                                   a
                                  1
                             0.8
                             0.6
                             0.4
                             0.2
                                  0                                                           t
                                  -6   -4      -2       0         2          4            6

                                                                  Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 9
2. Regularized big bang

Two possible scenarios:
 1. nonsingular bouncing cosmology [3, 4] from t = −∞ to t = ∞
    (valid for b ≫ lPlanck ?) [gravitational waves generated in the pre-
    bounce epoch keep on propagating into the postbounce epoch];
 2. new physics phase at t = 0 pair-produces [5] a “universe” for
    t > 0 and an “antiuniverse” for t < 0 (valid for b ∼ lPlanck ?).
For both scenarios, the t = 0 slice corresponds to a spacetime defect,
which manifests itself as a discontinuity of the extrinsic curvature K
on constant-t hypersurfaces. Also, there is a discontinuity at t = 0 of the
expansion function θ for a bunch of timelike geodesics [Wang, PhD thesis,
KIT, 2020].
It is not clear, for the first scenario, what physical mechanism determines
the relatively large value of b. For the second scenario, the hope is that
the new physics sets the value of b.                           ⇐ THIS TALK

                                              Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 10
3. New phase from M-theory

M-theory is a hypothetical theory that
unifies all five consistent versions of
10D superstring theory (cf. [6, 7]).

For an explicit description of the new phase replacing the big bang, we
use the IIB matrix model of Kawai and collaborators [8, 9], which has
been proposed as a nonperturbative definition of type-IIB superstring
theory (and, thereby, of M-theory).

                                             Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 11
3. New phase from M-theory

This IIB matrix model has N × N traceless Hermitian matrices, ten
bosonic matrices Aµ and essentially eight fermionic (Majorana–Weyl)
matrices Ψα .
The partition function Z of the Lorentzian IIB matrix model is defined
by the following “path” integral [8, 9, 10, 11]:
       Z                         Z
                            4
                                                 4
                                                     
  Z = dA dΨ exp i S/ℓ = dA exp i Seff /ℓ ,                           (9a)
                                                                   !
             1  µ ν  ρ σ                  1  e µ          ν  
   S = −Tr      A , A A , A ηeµρ ηeνσ + Ψβ Γβα ηeµν A , Ψα , (9b)
             4                                2
       h                     i
ηeµν = diag − 1, 1, . . . , 1    , for µ, ν ∈ {0, 1, . . . , 9} .    (9c)
                             µν
A model length scale “ℓ” has been introduced, so that Aµ has the
dimension of length and Ψα the dimension of (length)3/2 .
Expectation values of further observables will be discussed later.

                                              Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 12
3. New phase from M-theory

Now, the matrices Aµ and Ψα in (9a) are merely integration variables.
Moreover, there is no obvious small dimensionless parameter to
motivate a saddle-point approximation.

Hence, the conceptual question: where is the classical spacetime?

Recently, I have suggested to revisit an old idea, the large-N master
field of Witten [12], for a possible origin of classical spacetime in the
context of IIB matrix model [13].
First, I will remind you of this mysterious master field (name coined by
Coleman) and give you the final result.
Then, time permitting, I will highlight a few of the technical details
collected in the Appendices.

                                                Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 13
3. New phase from M-theory

Consider the gauge-invariant bosonic observable
                      µ1 ... µm       µ1      µm
                                                 
                    w           = Tr A . . . A     .                                                   (10)

Strings of these observables have expectation values
                                  Z
   µ1 ... µm   ν1 ... νn        1        µ1 ... µm   ν1 ... νn
                                                                     i S /ℓ4
hw           w           ···i =     dA w           w           · · · e eff . (11)
                                Z
The following factorization property holds to leading order in N :
                                        N
              h wµ1 ... µm wµ1 ... µm i = hwµ1 ... µm i hwµ1 ... µm i ,                                (12)

without sums over repeated indices. Similar large-N factorization
properties hold for all expectation values (11).
The leading-order equality (12) is a truly remarkable result for a
statistical (quantum) theory.

                                                      Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 14
3. New phase from M-theory

Indeed, according to Witten [12], the factorization (12) implies that
the path integrals (11) are saturated by a single configuration,
namely by the so-called master field A  b µ.

Considering one w observable for simplicity, we then have for its
expectation value (“Wilson loop”):
                                                                
                         µ1 ... µm    N       b µ1      b µm
                    hw               i = Tr A        ... A           ,                                    (13)

and similarly for the other expectation values (11).
Hence, we do not have to perform the path integrals on the right-hand
side of (11): we “only” need ten traceless Hermitian matrices Ab µ to get
all these expectation values from the simple recipe of replacing each
A µ in the observables by Ab µ , just as was done in (13).

                                                         Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 15
3. New phase from M-theory

Now, the meaning of the suggestion on slide 13 is clear:
                                                               b µ.
classical spacetime resides in the model master-field matrices A

Assume that the matrices A b µ of the Lorentzian-IIB-matrix-model
master field are known and that they are approximately band-diagonal.
Then, it is possible [13] to extract a discrete set of spacetime points {bxkµ }
and the emergent inverse metric g µν (x), with the metric gµν (x) obtained
as matrix inverse.
It is even possible [14] that the large-N master field of the Lorentzian
IIB matrix model gives rise to the regularized-big-bang metric (6) of GR.
Final result: effective length parameter b of the regularized-big-bang
metric (6) is calculated in terms of the IIB-matrix-model length scale ℓ,
                       ?          p
              beff ∼ ℓ ∼ lPlanck ≡ ~ G/c3 ≈ 1.62 × 10−35 m .            (14)

Technical details are collected in the Appendices.

                                                Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 16
4. Conclusions

It is conceivable that a new physics phase replaces the big bang
singularity suggested by our current theories (GR&SM).
For an explicit calculation, we have considered the IIB matrix model,
which has been proposed as a nonperturbative definition of type-IIB
superstring theory (M-theory).
The crucial insight is that the emergent classical spacetime may reside
in the large-N master field A  b µ of the model.
                                                b µ can give rise to the
In principle, the IIB-matrix-model master field A
regularized-big-bang metric with length parameter b ∼ ℓ, where ℓ is the
length scale of the matrix model.
At this moment, the outstanding task is to calculate the exact IIB-matrix-
model master field A           b µ or, at least, to get a reliable approximation of it...
.......................................................................................................................
Time permitting, we can mention a few technical details.                                       (→ slide 3)

                                                                       Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 17
5. References

 [1] C.W. Misner, K.S. Thorne, and J.A. Wheeler, Gravitation (Princeton University
     Press, Princeton, NJ, 2017).
 [2] F.R. Klinkhamer, “Regularized big bang singularity,” Phys. Rev. D 100, 023536
     (2019), arXiv:1903.10450.
 [3] F.R. Klinkhamer and Z.L. Wang, “Nonsingular bouncing cosmology from general
     relativity,” Phys. Rev. D 100, 083534 (2019), arXiv:1904.09961.
 [4] F.R. Klinkhamer and Z.L. Wang, “Nonsingular bouncing cosmology from general
     relativity: Scalar metric perturbations,” Phys. Rev. D 101, 064061 (2020),
     arXiv:1911.06173.
 [5] L. Boyle, K. Finn, and N. Turok, “CPT-symmetric Universe,” Phys. Rev. Lett. 121,
     251301 (2018), arXiv:1803.08928.
 [6] E. Witten, “String theory dynamics in various dimensions,” Nucl. Phys. B 443, 85
     (1995), arXiv:hep-th/9503124.
 [7] P. Horava and E. Witten, “Heterotic and type I string dynamics from eleven
     dimensions,” Nucl. Phys. B 460, 506 (1996), arXiv:hep-th/9510209.

                                                       Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 18
5. References

 [8] N. Ishibashi, H. Kawai, Y. Kitazawa, and A. Tsuchiya, “A large-N reduced model
     as superstring,” Nucl. Phys. B 498, 467 (1997), arXiv:hep-th/9612115.
 [9] H. Aoki, S. Iso, H. Kawai, Y. Kitazawa, A. Tsuchiya, and T. Tada, “IIB matrix
     model,” Prog. Theor. Phys. Suppl. 134, 47 (1999), arXiv:hep-th/9908038.
[10] S.W. Kim, J. Nishimura, and A. Tsuchiya, “Expanding (3+1)-dimensional universe
     from a Lorentzian matrix model for superstring theory in (9+1)-dimensions,”
     Phys. Rev. Lett. 108, 011601 (2012), arXiv:1108.1540.
[11] J. Nishimura and A. Tsuchiya, “Complex Langevin analysis of the space-time
     structure in the Lorentzian type IIB matrix model,” JHEP 1906, 077 (2019),
     arXiv:1904.05919.
[12] E. Witten, “The 1/N expansion in atomic and particle physics,” in G. ’t Hooft et. al
     (eds.), Recent Developments in Gauge Theories, Cargese 1979 (Plenum Press,
     New York, 1980).
[13] F.R. Klinkhamer, “IIB matrix model: Emergent spacetime from the master field,”
     to appear in Prog. Theor. Exp. Phys., arXiv:2007.08485.
[14] F.R. Klinkhamer, “Regularized big bang and IIB matrix model,” arXiv:2009.06525.

                                                        Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 19
A. Extraction of the spacetime points

Aoki et al. [9] have argued that the eigenvalues of the matrices A µ of
model (9) can be interpreted as spacetime coordinates, so that the
model has a ten-dimensional N = 2 spacetime supersymmetry.
Here, we will turn to the eigenvalues of the master-field matrices Ab µ.
Assume that the matrices A   b µ of the Lorentzian-IIB-matrix-model
master field are known and that they are approximately band-diagonal
(as suggested by numerical results [10, 11]). Then, make a particular
global gauge transformation [10],
                    µ
                  b
                  A     =    bµ Ω † ,
                            ΩA           Ω ∈ SU (N ) ,                                         (15)

so that the transformed 0-component matrix is diagonal and has
ordered eigenvalues α bi ∈ R,
            0                                         
          b
          A    = diag α  b1 , α
                              b2 , . . . , α
                                           bN −1 , α
                                                   bN ,                                      (16a)
                                                        N
                                                        X
          b1 ≤ α
          α        b2 ≤ . . . ≤ α    bN −1 ≤ α   bN ,     α
                                                          bi = 0 .                           (16b)
                                                          i=1
                                              Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 20
A. Extraction of the spacetime points

The ordering (16b) will turn out to be crucial for the time coordinate b
                                                                       t to
be obtained later.
A relatively simple procedure [13] approximates the eigenvalues of the
                    m
                  b
spatial matrices A but still manages to order them along the diagonal,
matching the temporal eigenvalues α  bi from (16).

We start from the following trivial observation:
if M is an N × N Hermitian matrix, then any n × n block centered on
the diagonal of M is also Hermitian, which holds for 1 ≤ n ≤ N .

Now, let K be an odd divisor of N , so that

                      N   =    K n,   K = 2L + 1,                                               (17)

where both L and n are positive integers.

                                               Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 21
A. Extraction of the spacetime points

                                        µ
                                      b
Consider, in each of the ten matrices A , the K blocks of size n × n
centered on the diagonal.
                                                0
                                              b
We already know the diagonalized blocks of A from (16a), which
allows us to define the following time coordinate b
                                                  t (σ) for σ ∈ (0, 1]:

                                          Xn
                                      1
               b0
               x              cb
                        k/K ≡ e t k/K ≡       α
                                              b(k−1) n+j ,                                        (18)
                                        n j=1

with k ∈ {1, . . . , K} and a velocity e
                                       c to be set to unity later. The time
coordinates from (18) are ordered,
                                                         
           b
           t 1/K ≤ b
                   t 2/K ≤ . . . ≤ b
                                   t 1 − 1/K ≤ b
                                               t 1 ,                                              (19)

because the α
            bi are, according to (16b).

                                                 Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 22
A. Extraction of the spacetime points

Next, obtain the eigenvalues of the n × n blocks of the nine spatial
                m                                         
             b                                        b m
matrices A and denote these real eigenvalues by β i , with
i ∈ {1, . . . , N }.

Define, just as for the time coordinate in (18), the following nine spatial
coordinates xb m (σ) for σ ∈ {(0, 1]:

                                 1 X h bm i
                                    n
                              
                    bm
                    x      k/K ≡       β             ,                                           (20)
                                 n j=1     (k−1) n+j

with k ∈ {1, . . . , K}.

                                                Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 23
A. Extraction of the spacetime points

                                 µ
                               b
If the master-field matrices A are approximately band-diagonal
and if the eigenvalues of the spatial n × n blocks show significant
scattering, then the expressions (18) and (20) may provide suitable
spacetime points, which, in a somewhat different notation, are denoted
                                                     
                µ
              bk = x
              x       bk0 , x
                            bkm ≡ x  b 0 k/K , x
                                               b m k/K ,            (21)

where k runs over {1, . . . , K}.
Each of these coordinates xbkµ has the dimension of length, which
traces back to the dimension of the bosonic matrix variable Aµ as
mentioned below (9c).

To summarize, with N = K n, the extracted spacetime points x  bkµ ,
for k ∈ {1, . . . , K}, are obtained as averaged eigenvalues of
the n × n blocks along the diagonals of the gauge-transformed
                            µ
                          b
master-field matrices A from (15)–(16).

                                             Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 24
B. Extraction of the spacetime metric

The points x bkµ effectively build a spacetime manifold with continuous
(interpolating) coordinates xµ if there is also an emerging metric gµν (x).
By considering the effective action of a low-energy scalar degree of
freedom σ “propagating” over the discrete spacetime points x    bkµ , the
following expression for the emergent inverse metric is obtained [9, 13]:
                Z
    g µν (x) ∼      dD y ρav (y) (x − y)µ (x − y)ν f (x − y) r(x, y) , (22a)
                  RD

   ρav (y)   ≡   hh ρ(y) ii ,                                                                 (22b)

with continuous spacetime coordinates xµ having the dimension of
length and spacetime dimension D = 9 + 1 = 10 for the original model.
The average hh ρ(y) ii corresponds, for the extraction procedure of App. A,
to averaging over different block sizes n and block positions along the
                                         µ
                                       b
diagonal in the master-field matrices A .

                                               Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 25
B. Extraction of the spacetime metric

The quantities that enter the integral (22) are the density function
                                XK
                                      (D)
                                                 
                       ρ(x) ≡       δ     x−x  bk ,                                                (23)
                                   k=1
the density correlation function r(x, y) defined by

               hh ρ(x) ρ(y) ii ≡ hh ρ(x) ii hh ρ(y) ii r(x, y) ,                                   (24)

and a localized real function f (x) from the scalar effective action,
                            X 1                      2
                 Seff [σ] ∼       f xbk − x
                                          bl σk − σl ,                                             (25)
                                2
                            k, l
where σk is the field value at the point x
                                         bk (the scalar degree of freedom σ
arises from a perturbation of the master field A b µ ; see App. A in Ref. [13]).
As r(x, y) is dimensionless and f (x) has dimension 1/(length)2 , the
inverse metric g µν (x) from (22) is seen to be dimensionless.
The metric gµν is simply obtained as the matrix inverse of g µν .

                                                  Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 26
B. Extraction of the spacetime metric

A few heuristic remarks [14] may help to clarify expression (22a).
In the standard continuum theory [i.e., a scalar field σ(x) propagating
over a given continuous spacetime manifold with metric gµν (x)],
two nearby points x′ and x′′ have approximately equal field values,
σ(x′ ) ∼ σ(x′′ ), and two distant points x′ and x′′′ generically have very
different field values, |σ(x′ ) − σ(x′′′ )|/|σ(x′ ) + σ(x′′′ )| & 1.
The logic is inverted for our discussion. Two approximately equal field
values, σ1 ∼ σ2 , may still have a relatively small action (25) if f (b
                                                                      x1 − x
                                                                           b2 )
∼ 1 and inserting f ∼ 1 in (22a) gives a “large” value for the inverse
metric g µν and, hence, a “small” value for the metric gµν , meaning that
the spacetime points x  b1 and x b2 are close (in units of ℓ).
Two different field values σ1 and σ3 have a small action (25) if f (b  x1 − x
                                                                            b3 )
∼ 0 and inserting f ∼ 0 in (22a) gives a “small” value for the inverse metric
g µν and, hence, a “large” value for the metric gµν , meaning that the space-
time points x b1 and x
                     b3 are separated by a large distance (in units of ℓ).

                                                Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 27
B. Extraction of the spacetime metric

To summarize, the emergent metric, in the context of the IIB matrix
model, is obtained from correlations of the extracted spacetime points
and the master-field perturbations.

The obvious question, now, is which spacetime and metric do we get?
We don’t know, as we do not have the IIB-matrix-model master field.
But, awaiting the final result on the master field, we can already
investigate what properties the master field would need to have in
order to be able to produce certain desired emerging metrics.
Some exploratory results are presented in App. C.

[Note that, in principle, the origin of the expression (22) need not be
the IIB matrix model but can be an entirely different theory, as long as
the emerging inverse metric is given by a multiple integral with the
same basic structure.]

                                              Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 28
C. Various emergent spacetimes

We restrict ourselves to four “large” spacetime dimensions [10, 11],
setting
                             D = 3 + 1 = 4,                          (26)

and use length units that normalize the Lorentzian-IIB-matrix-model
length scale,
                                ℓ = 1.                              (27)

Then, it is possible to choose appropriate functions ρav (y), f (x − y),
and r(x, y) in (22), so that the Minkowski metric is obtained [as given
by (2) for a2 (t) = 1].
Similarly, it is possible to choose appropriate functions ρav (y), f (x − y),
and r(x, y) in (22), so that the spatially flat Robertson–Walker metric
(2) is obtained.

                                                 Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 29
C. Various emergent spacetimes

In order to get an inverse metric whose component g 00 diverges at
t = 0, it is necessary to relax the convergence properties of the y 0
integral in (22a) by adapting the functions ρav (y), f (x − y), and r(x, y).
In this way, it is possible to obtain the following inverse metric [14]:
             2
            
               t + c−2
            
             −    2     ,                 for µ = ν = 0 ,
                 t
   µν
  g(eff)   ∼ 1 + c2 t2 + c4 t4 + . . . ,   for µ = ν = m ∈ {1, 2, 3} ,                              (28)
            
            
            
            
            
              0,                           otherwise ,

with real dimensionless coefficients cn that result from the requirement
that tn terms, for n > 0, vanish in g(eff)
                                     00
                                           .

                                                   Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 30
C. Various emergent spacetimes

The matrix inverse of (28) gives the following Lorentzian metric:
                   2
            
                  t
            
             − 2        ,                for µ = ν = 0 ,
            
             t + c−2
  (eff)                 1
 gµν      ∼                                                                                        (29)
            
                     2      4        ,   for µ = ν = m ∈ {1, 2, 3} ,
            
             1 + c2 t + c4 t + . . .
            
            0 ,                          otherwise ,

which has, for c−2 > 0, a vanishing determinant at t = 0 and is,
therefore, degenerate.

                                                  Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 31
C. Various emergent spacetimes

The emergent metric (29) has indeed the structure of the regularized-
big-bang metric (6a), with the following effective parameters:

                       b2eff   ∼   c−2 ℓ2 ,                                                         (30a)
                                                2
                   a2eff (t)   ∼   1 − c2 t/ℓ        + ... ,                                        (30b)

where the IIB-matrix-model length scale ℓ has been restored and
where the leading coefficients c−2 and c2 have been calculated [14].
By choosing the Ansatz parameters appropriately, we can get c2 < 0
in (30b), so that the emergent classical spacetime corresponds to the
spacetime of a nonsingular cosmic bounce at t = 0, as obtained in (8)
from Einstein’s gravitational field equation with a wM = 1/3 perfect
fluid.

                                                     Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 32
C. Various emergent spacetimes

The proper cosmological interpretation of the emergent classical
spacetime is perhaps as follows.
The new physics phase (replacing the big bang singularity of GR&SM)
is assumed to be described by the IIB matrix model and the corresponding
large-N master field gives rise to the points and metric of a classical
spacetime.
If the master field has an appropriate structure, the emergent metric
has a tamed big bang, with a metric similar to the regularized-big-bang
metric of GR [2] but now having an effective length parameter beff
proportional to the IIB-matrix-model length scale ℓ , as given by (30a).
In fact, one possible interpretation is that the new phase has produced
a universe-antiuniverse pair [5], that is, a “universe” for t > 0 and an
“antiuniverse” for t < 0.

                                              Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 33
D. More on the Lorentzian signature

Up till now, we have considered the Lorentzian IIB matrix model,
which has two characteristics:

 1. the “Lorentzian” coupling constants ηeµν from (9c);
                                      4
 2. the Feynman phase factor ei S/ℓ in the “path” integral (9a).

From the master field of this Lorentzian matrix model, we obtained
the spacetime points from expressions (18) and (20) in App. A and
the inverse metric from expression (22) in App. B.
Several Lorentzian inverse metrics were found in App. C, where the
Ansätze used [14] relied on having “Lorentzian” coupling constants ηeµν .
But there is another way [13] to obtain Lorentzian inverse metrics,
namely by making an appropriately odd Ansatz for the correlations
functions entering (22), so that the resulting matrix is off-diagonal.

                                               Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 34
D. More on the Lorentzian signature

With this appropriately odd Ansatz, it is, in principle, also possible
to get a Lorentzian inverse metric from the Euclidean matrix model,
which has nonnegative coupling constants δeµν in the action and a
                   4
weight factor e− S/ℓ in the path integral. The spacetime points are
extracted from the Euclidean master field (no gauge transformation
needed) by the expression (20), where m now runs over {1, . . . , D}.

The details of a toy-model calculation are as follows (expanding on
a parenthetical remark in the last paragraph of App. B in Ref. [13]).
The calculation starts from the multiple integral (22) for spacetime
dimension D = 4 by writing in the integrand
  f (x − y) r(x, y) = f (x − y) re(y − x) r(x, y) = h(y − x) r(x, y) ,                           (31)

where the new function r(x, y) has a more complicated dependence
on x and y than the combination x − y.

                                                Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 35
D. More on the Lorentzian signature

The D = 4 multiple integral (22), with y 0 replaced by y 4 , is then
evaluated at the spacetime point

                                        xµ = 0 ,                                                               (32a)

with the replacement (31) in the integrand and two further simplifications:

                     hh ρ(y) ii = 1 ,                  r(x, y) = 1 ,                                           (32b)

and symmetric cutoffs on the integrals,
                         Z   1                 Z   1
                                  dy 1 . . .           dy 4 .                                                  (32c)
                             −1                −1

The only nontrivial contribution to the integrand of (22) now comes
from the correlation function h as defined by (31).

                                                                Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 36
D. More on the Lorentzian signature

From (22) and (32), we then get the emergent inverse metric
                     Z   1           Z   1          Z   1           Z   1
     µν
    gtest,E4 (0) ∼            dy 1           dy 2            dy 3            dy 4 y µ y ν htest,E4 (y) ,                        (33)
                         −1          −1                 −1              −1

with the following Ansatz for the correlation function h :
                                1    2         1    3          1    4          2    3          2    4          3     4
                                                                                                                         
  htest,E4 (y) = 1 − γ y y + y y + y y + y y + y y + y y                                                                     , (34)

where γ multiplies monomials that are odd in two coordinates and even
in the two others.
Note that the Ansatz (34) treats all coordinates y 1 , y 2 , y 3 , and y 4 equally,
in line with the coupling constants δeµν of the Euclidean matrix model.

                                                                               Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 37
D. More on the Lorentzian signature

The integrals of (33) with Ansatz   function (34) are trivial and we obtain
                                                       
                             3       −γ    −γ     −γ
           µν        16   −γ        3    −γ     −γ   ,
         gγ (0) ∼                                                      (35a)
                      9  −γ         −γ     3     −γ 
                            −γ       −γ    −γ      3

where the matrix on the right-hand side has the following eigenvalues
and normalized eigenvectors:
        16 n                                        o
 Eγ =        (3 − 3 γ) , (3 + γ) , (3 + γ) , (3 + γ) ,             (35b)
        9
                                                      
       
              1               1                0          1    
                                                                
       1             1             1            1      
              1 , √        −1   , √        0  ,     1   . (35c)
 Vγ =                                              
       
         2    1         2     0         2      1      2   −1 
                                                               
               1               0                −1         −1

                                                Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 38
D. More on the Lorentzian signature

From (35b), we have the following signatures:
                            (+ − −−)          for γ ∈ (−∞, −3) ,                                                  (36a)

                            (+ + ++)          for γ ∈ (−3, 1) ,                                                   (36b)

                            (− + ++)          for γ ∈ (1, ∞) .                                                    (36c)

Hence, we obtain Lorentzian signatures for parameter values γ
sufficiently far away from zero, γ > 1 or γ < −3.
The conclusion is that it is, in principle, possible to get a Lorentzian
emergent inverse metric from the Euclidean IIB matrix model, provided
the correlation functions have the appropriate structure.
This observation, if applicable, would remove the need for working with
the (possibly more difficult) Lorentzian IIB matrix model.
———————————————————————————————-
Figure credits: en.wikipedia.org (Lady with an ermine); commons.wikimedia.org (M-theory sketch).

                                                                   Epiphany Conference — Cracow (Zoom), January 9, 2021 (v1.01) – p. 39
You can also read